Thanks, Matthew, for not only the answer but also the historical context and 
keyby.

Joe

-----Original Message-----
From: Matthew Dowle [mailto:[email protected]] On Behalf Of Matthew 
Dowle
Sent: Tuesday, April 10, 2012 8:22 PM
To: Joseph Voelkel
Cc: [email protected]
Subject: Re: [datatable-help] Order of DT after non-keyed by

Yes, correct behaviour. That's an ad hoc by which preserves the group order 
(the order of first appearance of each group). From ?data.table, the second 
part of this sentence :

"The order of the rows within each group is preserved, as is the order of the 
groups."

That was new in 1.6.3 :

o   Ad hoc grouping now returns results in the same order each 
    group first appears in the table, rather than sorting the
    groups. Thanks to Steve Lianoglou for highlighting. The order
    of the rows within each group always has and always will be 
    preserved. For larger datasets a 'keyed by' is still faster;
    e.g., by=key(DT).

To reorder the ad hoc by result, change 'by=' to 'keyby=' (in v1.8.0).
Not be confused with keyed by!

Matthew


On Tue, 2012-04-10 at 19:51 -0400, Joseph Voelkel wrote:
> Here is a simple example of a simple question:
> 
>  
> 
> dt<-data.table(a=rep(1:5,1:5),b=1,c=rep(1:3,5))
> 
> dt
> 
> dt[,seq_along(b),by=a] # expected behavior (note: dt is already in 
> order of a)
> 
> setkey(dt,c) # to sort by c
> 
> dt
> 
> dt[,seq_along(b),by=a] # expected behavior? Appears to be in order of
> unique(dt$a)
> 
>  
> 
> Thanks,
> 
>  
> 
> Joe Voelkel
> 
> 
> _______________________________________________
> datatable-help mailing list
> [email protected]
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable
> -help


_______________________________________________
datatable-help mailing list
[email protected]
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help

Reply via email to